Qwen3.8-Flash-Next

Alibaba llm

Open-weight MoE preview of the Qwen4 architecture (Gated DeltaNet + Qwen Sparse Attention); served on Qwen Cloud as Qwen3.8-Flash.

overall ⌄

61.2

32 of 125 ranked

price · $/M tokens

0.15 / 0.47

in / out, as of 2026-07-25

to run it yourself ⌄

108 GB

minimum · 432 GB recommended

Benchmarks

What its maker published, and what anyone else measured. Every figure links the document it came from.

benchmarkpublishedmeasured
GPQA91.7 —
SWE-bench Pro62.5 —
HLE35.9 —

Cheaper, and at least as good

On a 1,000-in, 1,000-out call these cost less and rank no lower — a hard comparison to argue with, and a narrow one. Value covers what it misses.

Claude Haiku 5.5 Anthropic$0.00060overall 65.3

Lineage

No recorded lineage — root or standalone model.

view in full graph →

Metadata

orgAlibaba
released2026-08-26
params180B
licenseother
hf idQwen/Qwen3.8-Flash-Next